• No results found

3. Results

3.2. Acute experiments

3.3.2. ROS formation H 2 DCFDA

Figure 16: Increase of average fluorescence measured for the probe H2DCFDA after 8 hours’ exposure of copepods to PS-NH2 particles. Each point on the X-axis refers to the separate concentrations measured with standard deviation lines for the average of the standard deviation of all the measurements. Y-axis indicates the average fluorescence measured.

Figure 16 shows the average raw fluorescence that was measured after the copepods had been exposed to PS-NH2 and the probe H2DCFDA for 8 hours. The concentrations between 1 µg/mL and 10 µg/mL of NPs show a lower fluorescence than the control. The concentrations 0.5 µg/mL and 25 µg/mL show higher fluorescence than the control, although this increase was not significant. Finally, copepods exposed to 50 µg/mL and 100 µg/mL of PS-NH2 had a significant increase compared to the control when the fluorescence was measured after 8 hours.

35 3.3.3. Lipid peroxidation C11-BODIPY581/591

Figure 17: Increase of average fluorescence measured for the probe C11-BODIPY after 8 hours’ exposure of copepods to PS-NH2 particles. Each point on the X-axis refers to the separate concentrations measured with standard deviation lines for the average of the standard deviation of all the measurements. Y-axis indicates the average fluorescence measured.

Figure 17 shows the average raw fluorescence from each concentration that was measured after the copepods had been exposed to PS-NH2 and the probe measuring lipid peroxidation for 8 hours. The concentrations from 0.5 µg/mL to 10 µg/mL and 50 µg/mL of NPs showed lower fluorescence than the fluorescence measured in the control. On the other hand, at 25 µg/mL a higher fluorescence was detected when compared with the control, although this increase was not significant (the standard deviation was extremely high). The highest concentration of PS-NH2 (100 µg/mL) displayed a significant increase in fluorescence when the copepods were exposed for 8 hours.

36

Figure 18: Fluorescence detected in copepods exposed to PS-NH2 and probe for lipid peroxidation, using the fluorescent microscope Olympus DP72. There are some elevations in fluorescence detected in the copepods, visible for 5 and 10 µg/mL.

In Figure 18, the fluorescent pictures of three different copepods co-exposed to PS-NH2 and C11-BODIPY581/591 show an increased fluorescence in areas associated with formation of lipid peroxidation. This increase in fluorescence seem to be comparable with the increase in fluorescence seen in Figure 15 for the DHR 123 probe. The fluorescence obtained for C11-BODIPY581/591 also seem to be mostly concentrated in the gut of the copepods, and may be caused by ingestion of particles. This possible ingestion of particles is most distinct for 5 µg/mL, and can be seen as light turquoise dots inside the copepod. The same dots can be seen for the 10 µg/mL exposure, only as larger areas in the rear region of the copepod. Additional fluorescent pictures taken for all concentrations can be found in Appendix B.

1 µg/mL 5 µg/mL 10 µg/mL

LPO

Bodipy

37

4. Discussion

The discussion is divided into 5 parts that cover 1) Characterisation of the NP particles, 2) Acute toxicity, 3) Oxidative stress, 4) Assessment of environmental relevance and 5) Evaluation of contribution to science and suggestions for future work.

4.1. Nanoplastic characterisation

As stated in literature, the toxicity of plastic particles may be influenced by their composition and properties, and will mainly depend on particle size, polymer type, surface charge and surface modifications (Besseling et al., 2014 (Bergami et al., 2016). For this reason, it is necessary to understand how NP particles characteristics change in different medias. The NP particle characterisation done through dynamic light scattering is shown in Table 3 in the results part 3.1, and shows that there are differences in how the particles behave in different media compositions. The NPs PMMA-COOH and PS-NH2 stayed approximately the same size in NSW compared to MQW, with an increase on 2 nm for PMMA-COOH and 75 nm for PS-NH2, respectively, as shown under Z-average in Table 3. This rather small increase in size for PS-NH2 was also observed by Bergami et al. (2017) and Manfra et al. (2017). In these studies, the particles size increased 69 nm and 54 nm from MQW to NSW, respectively. The other three polymer types (PMMA, PS and PS-COOH) formed micro-aggregates (all above 1500 nm) when exposed to NSW. This increase in Z-average was also observed by Bergami et al. (2017) and Manfra et al. (2017) for the particle PS-COOH, in which an increase of 1010 nm and 940 nm from MQW to NSW was recorded. These differences in aggregation were also confirmed visually for PMMA and PMMA-COOH. The pictures taken after the completion of the acute test (Figure 9 and 10) clearly demonstrate aggregation of PMMA after 48 hours exposure in NSW, while no aggregation was clearly evident for PMMA-COOH. The lack of visible particles means that the particles are probably too small to be detected under the microscope after 48 hours. Unfortunately, Z-average values for PMMA, PMMA-COOH and PS cannot be not found in other studies, and thus it was not possible to compare these sizes with others in the literature.

The determination of surface charge of NPs is important to understand particle behaviour, as the particle charge is known to affect their behaviour in terms of stability and aggregation (Bergami et al., 2017). All the particles showed a negative zeta-potential, the measure of surface charge, in MQW except for PS-NH2 which had a positive zeta-potential (Table 3). The NP PS-NH2 was the only particle in this study that was confirmed as positively

38

charged, as the zeta-potential was not measured in NSW for the non-functionalized and carboxylated particles due to stability issues during the measurements. However, in the study by Bergami et al. (2016), the PS-COOH (40 nm) particles used maintained a negative charge while suspended in seawater. Chen et al. (2018) also measured the zeta-potential for PS, PS-COOH, PMMA and PMMA-COOH (25 nm) in artificial seawater (ASW), and discovered a negative surface charge for all the particles. It is likely to assume that the zeta-potential for the particles would be similar in NSW, though naturally occurring colloids may affect their charge in NSW (Bergami et al., 2016).

The polydispersity index (PDI) measures the distribution of molecular mass in a polymer sample (far right in Table 3) (Rane & Choi, 20). A similar increase from MQW to NSW for PS-COOH and PS-NH2 (respectively an increase of 36 and 13) was found in other studies, and were proposed as being typical PDI values for these polymer types (Bergami et al., 2016; Manfra et al., 2017). Since the PDI measures distribution of molecular mass, it should correlate with Z-average. A more polydisperse sample (and therefore a high PDI) would reflect larger aggregates than a more monodisperse sample (Helseth & Ore, 2018a; Rane & Choi, 2005). There is a noticeable similar trend between the Z-average size and PDI measured for most of the polymer types. The polymer types PMMA-COOH and PS-NH2 with the smallest measured sizes in NSW (57.9 and 132.9 nm), were the least aggregated particles in NSW, and also showed the lowest increase in polydispersion. The polymer types PMMA and PS-COOH also followed the same pattern as the NPs already discussed, where PMMA had smaller aggregates than PS-COOH in NSW, and also the lowest PDI (Table 3). The NP polymer that did not follow this pattern was PS, which formed smaller aggregates than both PMMA and PS-COOH in NSW, but had the highest PDI value of all polymer types. This polymer type had a lower zeta-potential than all the other particles (-58.9 mV), which may have affected its PDI value. It is also possible that this polymer type behaved differently from the others because of the lack of functional groups. The PS has a higher zeta-potential than PS-COOH when measured in artificial seawater (ASW) (-20 and -21 mV, respectively), which can indicate that natural organic matter may have interfered and altered the properties of these NPs (Bergami et al., 2016; Chen et al., 2018).

The difference in sizes of aggregates observed in Table 3 may occur because of differences in stability for each of the polymer types. NSW has a high ionic strength, and may screen the particle surface charges of the polymers, causing the visible aggregation (Bergami et al., 2016). The NP PS-COOH was negatively surface charged, while PS-NH2 had a positive surface charge, which may answer why agglomeration was observed for only PS-COOH

39 (Bergami et al., 2017). An interesting observation concerning the agglomeration, is that one can observe that the polymer type must have something to say for the agglomeration of particles, as well as the presence of functional groups. This is evident when comparing the polymer types PMMA-COOH and PS-COOH, which are made of different polymer cores, but have the same functional groups. The NP PMMA-COOH stayed approximately the same size in NSW as in MQW, while PS-COOH originated micro-aggregates in NSW (Table 3). The NPs PMMA-COOH and PS-PMMA-COOH have similar zeta-potentials in MQW, but since the zeta-potential was unfortunately not measured in NSW, it is uncertain if the surface charge of the particles was causing aggregation of one polymer type while the other stayed monodispersed. Booth et al.

(2016) suggested that the visible agglomeration for all particle agglomerating NP polymers may occur because of heteroaggregation with other particles in the natural seawater. Even though the water used in this thesis was filtered at 0.22µm, it is possible that other parameters than the ones measured may alter NPs characteristics, like natural organic matter (NOM), proteins and ionic salts, that are not present in MQW. Because of the complexity of the aquatic environment, there are many factors that may trigger the observed aggregation, but it is not easy to say exactly what the cause is.

40

4.2. Acute test

Very few NPs have been properly tested for toxicity, and of the existing studies very few show lethal endpoints and corresponding EC50 values (Walker et al., 2012). There are no studies that report the role of particle type and functionalisation on toxicity to the marine copepod T.

battagliai in literature, being this study the first one to look at the effects of plain PS and PMMA in comparison to carboxylated and aminated PMMA and PS. The results obtained showed large variances in toxicity for all particles, from EC50 values of about 7.8 µg/mL for PS-NH2 to above 100 µg/mL for PMMA. These results clearly demonstrated that polymer type, size, functional groups and surface charge of the particles affected their toxicity. It was for example seen that PMMA-COOH (EC50 = 89.5 µg/mL) and PS-COOH (EC50 ≈ 69.3 µg/mL) were apparently more toxic than the plain polymers PMMA (EC50 >100 µg/mL) and PS (EC50 ≈ 95.6 µg/mL).

A reason why polymers with functional groups seem more lethal than plain polymers, is possibly because the functional groups make the molecular structure of the plastic particles similar to the structure of proteins. Such a structure will make these particles cross easier over cell membranes than the similar particles without functional groups (Bergami et al., 2017).

In another study, microcrustaceans (including D. magna and Corophium volutator, a marine crustacean) were exposed to non-carboxylated PMMA (125 nm) (Booth et al. (2016). With an EC50 value of >1000 µg/mL for D. magna and >500 µg/mL for C. volutator, this particle type was found not to cause acute toxicity in these crustaceans at reasonable concentrations (Booth et al., 2016). The findings made by Booth et al. (2016) cohere with the results obtained in this thesis, where an EC50 value >100 µg/mL was found for PMMA. Studies on PMMA-COOH toxicity do not exist yet, so direct comparison between effect values could not be conducted. In this thesis, there was a slightly higher mortality for the carboxylic polymer PMMA-COOH than for the non-functionalised PMMA, (Figure 12 a, b). When reviewing the z-average sizes of these particles, it is seen that PMMA agglomerated in NSW while PMMA-COOH stayed approximately the same size. Since PMMA-COOH remained the same size over a longer period of time, the plastic particle was possibly more toxic to the copepods, due to larger potential for ingestion and interaction with cells and intracellular biological targets (Bergami et al., 2017).

When studying the results for the PMMA particles, it is likely to believe that it is the functional –COOH group that made the polymer stable in NSW and not agglomerate. These results further confirm that the functional group also makes the particle more toxic, as is more similar to proteins, as discussed above (Bergami et al., 2017).

41 Bergami et al. (2016) tested the acute toxicity of the particle PS-COOH and PS-NH2, but with another life stage of another marine crustacean, the larvae stadium of the brine shrimp.

The results authors got cohere with the results obtained in this thesis, as well as the results showed by Della Torre et al. (2014) with sea urchins embryos, where the PS-NH2 appears to be more lethal than the PS-COOH particles. The results for PS-COOH and PS-NH2 found in the two abovementioned articles coheres with the findings in this thesis, as the results here show EC50 values of about 69.3 and 7.8 µg/mL, respectively. Della Torre et al. (2014) did not find an EC50 for PS-COOH, as this particle did not show any relevant effects on embryo development of sea urchins, while the EC50 for PS-NH2 obtained was 3.8 µg/mL (Della Torre et al., 2014).

It must be emphasised that the EC50 values found by these authors cannot be directly compared with the lethal concentrations found in this thesis, as their EC50 values show the concentration that affects embryo development, not the concentration that kills them. Embryo development and adult mortality are fundamentally different endpoints and may be caused by dissimilar toxic mechanisms. However, the results found from Della Torre et al. (2014) showed the same trend as the EC50 values found in this thesis, and verified that PS-NH2 was more toxic than PS-COOH.

The NP PS without any functional groups had a smaller size in NSW than PS-COOH, (Table 3), but displayed higher mortality (EC50 ≈ 69.2 µg/mL) than non-functionalised PS (EC50

≈ 95.6 µg/mL) (Table 5). This contrasted with what was observed for PMMA and PMMA-COOH, where the smallest particle size displayed the largest toxic potency in terms of mortality. These variances may indicate that the functional groups are the main reason for why the copepods react differently when being exposed to the particles, potentially due to their difference in surface charge (Bergami et al., 2017; Della Torre et al., 2014). The most lethal polymer type observed in this study was without doubt PS-NH2 (EC50 ≈ 7.8 µg/mL). The NP PS-NH2 was the only plastic particle that was lethal for the crustaceans showing a clear concentration-dependent increase in mortality. This was also the only particle with a positive surface charge, as seen in Table 3. Della Torre et al. (2014) stated that both COOH and PS-NH2 entered the cells of the tested organisms, although it was only PS-NH2 that caused lipid peroxidation and ROS formation. The authors suggest that it is the positive surface charge that makes the PS-NH2 particles so toxic.

42

4.3. Oxidative stress determination

Several studies have suggested that particle surface chemistry is relevant for the toxic potential of NPs, not only for cell death and/or death of the organism, but also for sub-lethal toxicity (Bergami et al., 2017). The results obtained in the acute testing showed that PS-NH2 was the only plastic polymer that was considered toxic from the polymers tested, and this acute toxicity seem to be associated with the type of particle functionalisation, as well as it surface charge.

With mortality seen at low NP concentrations, this was the polymer type tested with the largest toxic potential, and it is interesting to try to understand how this polymer affected the organisms on a sub-lethal level.

The formation of oxidative stress is one example of a sub-lethal endpoint that can be determined in aquatic organisms, and it has been indicated as one of the possible effects of NP particles (Della Torre et al., 2014; Jeong et al., 2018). In the present study, for the two ROS-detecting probes, DHR123 and H2DCFDA, it seems like H2DCFDA was the most suitable to determine ROS production after exposure to PS-NH2. With significant fluorescence for the two highest concentrations (50 µg/mL and 100 µg/mL) (Figure 16), this probe seems the most applicable when detecting ROS formation in copepods. This probe was also the only that showed a significant fluorescence after the normalisation of data for 100 µg/mL (Figure B19), which indicates the formation of ROS, although this method needs to be further optimised to get better results. Regular ROS formation after exposure to NPs was also detected by Jeong et al. (2018) using the fluorescent probe H2DCFDA. In this study, rotifers (which are about the same size as copepods) were exposed to non-functionalised PS, and significant differences were found at concentrations as low as 0.1 µg/mL in the tested animals. However, in the study by Jeong et al. (2018) another method was used for the detection of ROS; about 2000 rotifers were exposed to the NPs for 24 hours, washed and then homogenized. The supernatant was used in this case for the probe test to detect ROS formation and lipid peroxidation instead of an in vivo method like the one used in this thesis. Artificial seawater was also used by Jeong et al. (2018) instead of natural seawater, as in this study. Even though the natural seawater used in this thesis is filtered at 0.22 µm, it is possible that the water contained other substrates (like NOM) that may have affected the fluorescent probes or the NPs, and be the reason for the elevated fluorescence seen for the blank controls.

One of the most common examples of physiological damages associated with oxidative stress is the formation of lipid peroxidation (Mylonas & Kouretas, 1999). The study from Jeong et al. (2018) imply that NP particles cause harm to membranes in chemical or physical ways.

43 In this thesis, the fluorescent probe C11-BODIPY was used to see if NP cause lipid peroxidation in T. battagliai. Cheloni and Slaveykova (2013) found the probe C11-BODIPY suitable for in vivo measurements of lipid peroxidation within green alga in MQW, and this is the first time this probe was used in this copepod species. Lipid peroxidation was one of the causes of membrane damage in rotifers exposed to PS NPs, and was detected by increased levels of malondialdehyde (MDA) which is the final product of lipid peroxidation (Jeong et al., 2018).

In this study, even though the results obtained for C11-BODIPY after normalisation (Appendix B, Figure B17) showed no significant increase in fluorescence, and therefore no lipid peroxidation, the probe gave an increase in signal at the highest concentration tested (Figure 16), and should be considered in further studies after additional method development.

For both lipid peroxidation and ROS formation, Jeong et al. (2018) observed a decrease in fluorescence for the highest concentration tested (20 µg/mL). This finding does not correlate with what is seen in this thesis, where only the highest concentration had an increase in fluorescence. This could occur because of differences in the composition of water, properties of the NPs in the water, use of different species or life stage differences. These differences in results can also occur because of the dissimilarities in the methods used, as the current study used in vivo methods and the study of Jeong et al. (2018) used another method that is already described. Even though the results are not completely comparative, both studies see a significant increase in fluorescence on some levels, indicating that NP particles seem to induce oxidative stress in the organisms.

For both lipid peroxidation and ROS formation, Jeong et al. (2018) observed a decrease in fluorescence for the highest concentration tested (20 µg/mL). This finding does not correlate with what is seen in this thesis, where only the highest concentration had an increase in fluorescence. This could occur because of differences in the composition of water, properties of the NPs in the water, use of different species or life stage differences. These differences in results can also occur because of the dissimilarities in the methods used, as the current study used in vivo methods and the study of Jeong et al. (2018) used another method that is already described. Even though the results are not completely comparative, both studies see a significant increase in fluorescence on some levels, indicating that NP particles seem to induce oxidative stress in the organisms.